Counting Iron-Absorbed Small Intestinal CellsBackgroundProject OverviewPreliminary Image ProcessingImage to be used.4 Test cases.Test Case 1 and Initial ProcessingMore Successful AttemptsCase 2Case 3Case 4Next step.Higher level obstaclesConclusionsCounting Iron-Absorbed Small Intestinal CellsJoe HalfenBackgroundThe counting of the cells is done by hand.Previous work tried to use statistics to estimate the number of cells on a slide.Would like to use more reliable techniques to determine the number of cellsProject OverviewWill use 4 test cases of a slide.Try to determine the best threshold and morphological techniques to accurately determine cell count.Give preliminary estimates on cell counts for the technique decided upon.Preliminary Image ProcessingExploit the color of the cellsSubjective trial and error processUsing threshold on the RGB colors before converting to gray map to do final threshold.Image to be used.4 Test cases.Best case ImageImage with additional region boundariesImage with no cellsImage with cells overlappingTest Case 1 and Initial ProcessingOriginal ImageInitial TryMore Successful AttemptsCase 2Case 3Case 4Next step.Image processing to determine region boundaries.Fill in regions.Threshold to binary image.Count cells.Higher level obstaclesSeparating cellsWatershed Method.Statistics from image.Statistics from user input.ConclusionsUse Matlab implementation.Determine good method for preliminary processing.Identify addition adjustments that will need to be made to the program to make it most
View Full Document